Introduction
Otoacoustic emissions (OAEs) are audio frequencies that are transmitted from the cochlea to the middle ear and into the external ear canal. The OAEs are elicited spontaneously or in response to stimulation. Distortion product OAEs (DPOAEs) are a type of OAE, which are recorded by simultaneously presenting pure tones of two different frequencies (f1 and f2). In response to two pure tones, a healthy cochlea generates additional frequencies that are arithmetically related to those pure tones which can be recorded as OAEs in the external ear canal. The most robust DPOAE response is located in the 2f1-f2 frequency.
Clinically, DPOAEs are recorded across different f1 and f2 frequencies for fixed intensity levels, referred to as a distortion product gram (DP-gram). In contrast, DPOAE input–output functions measure DPOAE amplitude as a function of stimulus intensity for one fixed frequency. The DPOAE input–output functions across different frequencies provide more comprehensive information about how OAEs change with intensity and frequency of the primary tones. The DPOAEs can be recorded at conventional frequencies with multiple sets of stimulus frequencies f1 and f2 between 0.5 and 8 kHz. Given the advancement of instrumentation, DPOAEs can be measured as high as 16 kHz and are as reliable as those measured at conventional frequencies. The measurements are repeatable in individuals with normal hearing and in clinical populations.Reference Dreisbach and Siegel1–Reference Dreisbach, Zettner, Chang Liu, Meuel Fernhoff, MacPhee and Boothroyd3
Clinically, DPOAEs are recorded with an f2/f1 ratio equal to 1.22, using the two pure tones of L1 = 65 dB SPL and L2 = 55 dB SPL. The f2/f1 ratio of 1.20–1.22 is most effective in the majority of patients, and frequency values are used as a default value in clinical DPOAE instrumentation. Optimum DPOAE amplitudes are measured when the L1 is higher than L2 by 10–15 dB.Reference Dhar and Hall4 The DPOAE input–output function is obtained by measuring the change in DPOAE level, with a systematic change in stimulus intensity (L1 and L2) between 30 and 70 dB SPL, while the frequency is held constant. The DPOAE input–output function provides a wealth of information about DPOAE amplitude as a function of different stimulus intensity levels.
Distortion product OAEs are commonly used as a diagnostic tool in audiological assessments of different age groups. Successful OAEs recording is not related to cochlear status alone and is significantly influenced by middle-ear conditions too.Reference Keefe, Gorga, Neely, Zhao and Vohr5 Detection and valid measurement of DPOAEs are influenced by middle-ear status, which may affect the forward transmission of stimuli and reverse transmission of OAEs. Abnormal or absent DPOAEs measured at conventional frequencies are reported in the case of various middle-ear conditions such as otitis media with effusion (OME), chronic suppurative otitis media (CSOM), tympanic membrane perforation and negative middle-ear pressure.Reference Yeo, Park, Park and Suh6–Reference Thompson, Henin and Long9
Alterations in terms of reduced DPOAE responses have also been reported in individuals with a previous history of OME.Reference De P Campos, Sanches, Hatzopoulos, Carvallo, Kochanek and Skarżynski10 A long-standing middle-ear condition might cause minor, but irreversible damage to the middle ear or cochlea. Otoacoustic emissions can be used as a tool to measure this subclinical damage.
Distortion product OAEs at ultra-high frequencies have been studied in individuals with normal hearing and in a variety of clinical populations. Distortion product OAEs recorded at ultra-high frequencies were found to be as reliable as the DPOAEs recorded at a conventional frequency range. Distortion product OAEs of more than 8 kHz were more variable than those recorded at low frequencies, and DPOAEs at high frequencies were found to be repeatable.Reference Dreisbach, Long and Lees2 Distortion product OAEs have also been found to be repeatable in patients with cystic fibrosis.Reference Dreisbach, Zettner, Chang Liu, Meuel Fernhoff, MacPhee and Boothroyd3 The biological inner-ear processes involved in the generation of OAEs at ultra-high frequencies are similar to those responsible for the generation of OAEs at a conventional frequency range.Reference Dreisbach and Siegel1
Distortion product OAEs at ultra-high frequencies were studied in children with and without middle-ear effusion.Reference Kei, Brazel, Crebbin, Richards and Willeston11 The findings revealed that the effect of middle-ear effusion in the ultra-high frequency region was not as severe as that in the lower frequency region. In that study, all subjects had a hearing threshold within 20 dB HL from 0.5 to 8 kHz, and DPOAEs were measured only at one level of the primary tones. The DPOAE input–output function provides a wealth of information about DPOAE amplitude as a function of different stimulus intensity levels. Most studies investigating DPOAEs in individuals with middle-ear disorders reported a decrease in OAE amplitude or a complete absence of the response in conventional frequencies.Reference Kemp, Ryan and Bray12
The influence of middle-ear conditions on OAEs is more complex. There is a need to understand the effect of middle-ear effusion on ultra-high frequencies for a meaningful interpretation of the OAE findings in these populations. There is a dearth of information on the influence of middle-ear conditions on ultra-high frequency DPOAEs. There is a need to understand the use of ultra-high frequency DPOAEs in cases of abnormal DPOAE findings for conventional frequencies related to the middle-ear condition.
Materials and methods
Participants
This study involved a control group, comprising 37 adults (15 males and 22 females) with normal hearing sensitivity, and a clinical group, consisting of 37 adults (15 males and 22 females) with CSOM. Participants in both groups were aged 19–35 years. The mean age of participants in the control group was 25.83 years (range = 19–35 years; standard deviation (SD) = 4.33) and the mean age of participants in the clinical group was 27.92 years (range = 19–35 years; SD = 5.53). The age and gender of participants were matched between the groups.
The participants in the control group fulfilled the inclusion criteria of: no history of otological disease, normal tympanometry findings and acoustic reflex measurements, normal hearing sensitivity (i.e. hearing thresholds within 15 dB HL for 0.5, 1, 2 and 4 kHz on pure tone audiometry), and no history of hazardous noise exposure or ototoxic medication.
In the clinical group, participants with middle-ear disorders, who had hearing thresholds of up to 55 dB HL due to CSOM, were included in the study. All the participants in the clinical group had chronic middle-ear inflammation, a tympanic membrane perforation and chronic ear discharge persisting for at least two weeks.13 The diagnosis of CSOM was made via a detailed case history, otoscopic examination and audiological evaluation.
Otoscopic examination revealed normal findings in the control group, and perforation of the tympanic membrane and purulent ear discharge in the clinical group. All participants in the clinical group had active middle-ear effusion.
The mean (and SD) of air conduction hearing thresholds from 0.25 to 16 kHz, for participants in both groups, are shown in Figure 1. The bone conduction hearing thresholds from 0.25 to 4 kHz were within normal limits (i.e. hearing thresholds were within 15 dB HL for 0.5, 1, 2 and 4 kHz for all participants in both groups). Hearing thresholds within 25 dB HL were considered as normal in ultra-high frequencies.
Fig. 1. Mean air conduction thresholds in the control and clinical groups. Error bars indicate 2 standard deviations from the mean.
Immittance evaluation showed that all participants in the control group had a type ‘A’ tympanogram, with a mean ear canal volume of 1.75 ml (range = 1.20–2.30 ml; SD = 0.28), and acoustic reflexes were present at 0.5, 1, 2 and 4 kHz. All participants in the clinical group had a type ‘B’ or flat tympanogram, with a mean ear canal volume of 1.80 ml (range = 1.10–2.40 ml; SD = 0.34), and absent acoustic reflexes for ipsilateral and contralateral stimulations.
Otoscopic examination, pure tone audiometry, immittance evaluation and DPOAE measurements were performed during the same session for all participants.
Procedure
All assessments were carried out in a sound-treated room, with appropriate acoustic isolation and noise within maximum permissible levels.
Measurement of pure tone thresholds
A calibrated Inventis Piano two-channel diagnostic audiometer (Padova, Italy) was used for estimating the hearing thresholds of all participants, using the modified Hughson and Westlake procedure. TDH-39 headphones were used for estimating air conduction thresholds at octave frequencies 0.25, 0.5, 1, 2, 4 and 8 kHz. HDA 300 headphones were used to estimate air conduction thresholds at 9, 10, 12.5 and 16 kHz. The B-71 bone vibrator was used for estimating bone conduction thresholds at octave frequencies 0.25, 0.5, 1, 2 and 4 kHz. The audiometer used for threshold estimation met the American National Standards Institute (‘ANSI S3.6-2010’) Specification for Audiometers standard.
Immittance evaluation
A calibrated GSI Tympstar Pro™ clinical tympanometer was used for immittance measurement. Tympanometry, and ipsilateral and contralateral acoustic reflexes for pure tones at 0.5, 1, 2 and 4 kHz, were measured for all participants. The probe tone frequency of 226 Hz at 85 dB SPL was used for all immittance measurements. Ear canal pressure was varied from + 200 to −400 daPa at the pressure sweep rate of 600 daPa/second during the tympanometry measurements. The immittance equipment used in the study met the American National Standards Institute (‘ANSI S3.39-1987 (R 2012)’) Specifications for Instruments to Measure Aural Acoustic Impedance and Admittance standard.
Distortion product otoacoustic emission input–output functions
Distortion product OAE (DPOAE) input–output functions were recorded at conventional frequencies and ultra-high frequencies. A calibrated Mimosa Acoustics/Starkey DP2000 OAE device was used for recording DPOAE input–output function. Appropriate probe fitting was ensured, and the stimulus was calibrated each time before recording DPOAEs, using the in-ear calibration method.
After fitting the probe tip in the ear canal, the calibration was performed by presenting the chirp stimulus in each of two channels (transducers), with a pause between the stimuli. The pressure frequency response in the ear canal was measured and plotted in the graph. The overlap of two calibration curves on each other (reproducibility of 95 per cent or more) was considered as good calibration and probe fitting. If the reproducibility of two curves was below 95 per cent, the probe was removed and fitted again for repeat calibration until a proper probe fit was obtained. The reproducibility in the current study was 95 per cent or more between the two channels for all DPOAE measurements.
If the primary tone levels deviated from actual sound pressure levels, the test was aborted. Subsequently, the probe was re-fitted, recalibration was performed and DPOAEs were re-measured.
Distortion product OAEs at conventional frequencies and ultra-high frequencies were recorded at the f2/f1 ratio of 1.22, with a systematically varying intensity of f1 (L1) or f2 (L2) in 5 dB steps, between 30 dB SPL and 70 dB SPL. The intensity of f1 (L1) was 10 dB higher compared to the intensity of f2 (L2) in all the DPOAE measurements. Distortion product OAEs were measured across f2 frequencies of 0.5, 1, 2, 4, 8, 9, 10.25, 12.5, 14 and 16 kHz. Distortion product OAE amplitudes and signal-to-noise ratios were estimated for all the participants across 7 intensity levels, at 10 different frequencies. Distortion product OAEs were considered measurable if the amplitude was 6 dB or above the noise floor (signal-to-noise ratio of 6 dB or more), with a minimum DPOAE amplitude of −20 dB SPL.
Results
All statistical analyses were performed using SPSS Statistics for Windows, version 20.0 (IBM, Armonk, New York, USA). The independent sample t-test showed no significant difference in age between the control group and the clinical group (p > 0.05). The one-sample Kolmogorov–Smirnov test was performed to assess the normal distribution of DPOAE signal-to-noise ratios (Appendix 1) and DPOAE amplitude data (Appendix 2) in both groups. The results showed that all the DPOAE amplitude and DPOAE signal-to-noise ratio data followed a normal distribution (p > 0.05) in both groups. Distortion product OAE (DPOAE) amplitudes and signal-to-noise ratios across all intensities and frequencies were compared between the groups using the independent sample t-test. The chi-square and Fisher's exact tests were used to measure the association between the presence or absence of middle-ear disorders and the rate of ears with measurable DPOAEs of 6 dB or more signal-to-noise ratio with a minimum DPOAE amplitude of −20 dB SPL.
The independent sample t-test was administered to examine the difference in mean DPOAE amplitudes between the groups across 7 L1 intensities and 10 f2 frequencies. The results revealed no significant differences in DPOAE amplitudes between the groups at L1 intensities of 40, 45 and 50 dB SPL, across all frequencies. At conventional frequencies, significant differences in DPOAE amplitudes were observed at 0.5, 1, 2 and 4 kHz between the groups. However, at ultra-high frequencies, significant differences in DPOAE amplitudes between the groups were observed only at 10.2 kHz. Significant differences were identified in DPOAE amplitudes among the two groups at: L1 intensities of 60, 65 and 70 dB SPL at 0.5, 2 and 10.2 kHz; L1 intensities of 55, 60, 65 and 70 dB SPL at 1 kHz; and L1 intensity of 65 dB SPL at 4 kHz.
Figure 2 depicts the mean DPOAE input–output functions across 10 different f2 frequencies, from 0.5 to 16 kHz, for L1 intensities 40, 45, 50, 55, 60, 65 and 70 dB SPL. The mean noise floor was similar in both groups. The noise floor values ranged from −5 to −25 dB SPL at 0.5 kHz, and from −10 to −30 dB SPL across all other test frequencies.
Fig. 2. Mean distortion product otoacoustic emission (DPOAE) input–output functions in the control and clinical groups at conventional and ultra-high frequencies. Error bars indicate 1 standard deviation from the means. Asterisks indicate significant differences between the two groups (p < 0.05).
Similar to the DPOAE amplitudes, the mean signal-to-noise ratios were higher for the control group compared to the clinical group (Figure 3). Significant differences in DPOAE signal-to-noise ratios (p < 0.05) were identified between groups at: L1 intensities of 65 and 70 dB SPL at 0.5 kHz; L1 intensities of 60, 65 and 70 dB SPL at 1 kHz; L1 intensities of 70 dB SPL at 2 and 8 kHz; and L1 intensities of 55, 60, 65 and 70 dB SPL at 4 kHz. Unlike conventional test frequencies, mean DPOAE signal-to-noise ratios at ultra-high frequencies were fairly similar for both groups. However, there were significant differences between groups in DPOAE signal-to-noise ratios (p < 0.05) at: L1 intensity of 65 dB SPL at 10.2 kHz; and L1 intensity of 60 dB SPL at 14 kHz. The other test frequencies revealed no significant differences in DPOAE signal-to-noise ratios between the two groups.
Fig. 3. Mean distortion product otoacoustic emission (DPOAE) signal-to-noise ratio (SNR) values in the control and clinical groups at conventional and ultra-high frequencies. Error bars indicate 1 standard deviation from the means. Asterisks indicate significant differences between the two groups (p < 0.05).
The chi-square and Fisher's exact tests were used to determine whether there was any association between the two groups in terms of the presence or absence of measurable DPOAEs (i.e. minimum signal-to-noise ratio of 6 dB or greater across all tested frequencies). The minimum DPOAE amplitude of –20 dB SPL was considered during analysis; hence, the measurable DPOAEs comprise ‘present DPOAEs’ and ‘present and abnormal DPOAEs’, given the low amplitude observed in the clinical group. This association was measured for L1 intensity from 55, 60, 65 and 70 dB SPL across all tested frequencies, given that the majority of participants in the clinical group had no measurable DPOAEs below the L1 intensity of 55 dB SPL.
The results of the chi-square and Fisher's exact tests revealed a significant difference in the rate of ears with measurable DPOAEs between the control group and the clinical group, as shown in Table 1. Significant differences were found in L1 intensity of: 65 and 70 dB SPL at 0.5 kHz; 55–70 dB SPL at 1, 2 and 4 kHz; 65 and 70 dB SPL at 8 kHz; and 60 SPL at 14 kHz.
Table 1. Rates of measurable DPOAEs for L1 intensities of 55–70 dB SPL for all tested frequencies in control and clinical groups*
Data represent numbers (and percentages) of ears with measurable DPOAEs.
* Total ears, n = 74 (37 ears in each group).
†Indicates significance (p < 0.05) on chi-square or Fisher's exact test. DPOAE = distortion product otoacoustic emissions
Discussion
The present study investigated the effect of CSOM on the DPOAE input–output functions for conventional and ultra-high frequencies. The results showed a significant reduction in DPOAE amplitudes at higher intensity levels for 0.5, 1, 2, 4 and 10.2 kHz. These significant differences between the groups were observed at higher intensity levels beyond the L1 intensity of 55 dB SPL. Significant reductions in DPOAE signal-to-noise ratios at the conventional frequency range were observed from 0.5 to 4 kHz at higher intensity primary tones in the clinical group. In the ultra-high frequency range, a significant difference in DPOAE signal-to-noise ratio was observed at one intensity level at 10.2 and 14 kHz.
A study of young adults with otitis media with effusion (OME) revealed a significant reduction in DPOAE amplitude in all tested frequencies from 0.75 to 8 kHz.Reference Yilmaz, Karasalihoglu, Tas, Yagiz and Tas14 A significant reduction in DPOAE signal-to-noise ratio was observed between 1 and 8 kHz in the clinical group in the present study. However, a greater reduction in DPOAE parameters was observed in the clinical group for frequencies of 4 kHz or less, as compared to higher frequencies. In one study, the DPOAE parameters were significantly different in children with OME, especially in the low frequencies.Reference Akdogan and Özkan15 Kei et al. studied DPOAEs in conventional and ultra-high frequencies in children with and without middle-ear dysfunction.Reference Kei, Brazel, Crebbin, Richards and Willeston11 Their results showed a significant reduction in DPOAE amplitude and signal-to-noise ratio at 1, 1.5, 2, 3, 4, 6, 8 and 13 kHz for the ‘fail immittance’ group when compared to the ‘pass immittance’ group.
The present study finds support from the above-mentioned studies suggesting that DPOAE parameters are significantly affected in individuals with CSOM at conventional frequencies as compared to ultra-high frequencies. The overall DPOAE amplitude measured in the study by Kei et al. was higher than that reported in the present study for both the control and clinical groups.Reference Kei, Brazel, Crebbin, Richards and Willeston11 The differences in DPOAE amplitude between the two studies could be attributed to the differences in participants’ age ranges in the two studies; that is, children versus adult participants.
The effect of CSOM is more pronounced on DPOAEs in the conventional frequency range compared to the smaller effect seen in the ultra-high frequency range. (Appendix 3 shows some screenshots of DPOAE recordings at ultra-high frequencies in the clinical group). A possible explanation for better DPOAE amplitude at ultra-high frequencies is increased middle-ear stiffness, because of negative pressure, fluid build-up and structural changes due to chronic otitis media. This increased stiffness could improve the tympanic membrane vibration at high frequencies.Reference Dai, Cheng, Wood and Gan16 Otitis media could affect the OAEs both by reducing the transmission of emission levels from the cochlea through the middle ear and by attenuating the stimulus reaching the cochlea.Reference Saleem, Ramachandran, Ramamurthy and Kay17 A previous study reported that middle-ear effusion reduces the number of measurable responses and their DPOAE amplitudes.Reference Topolska, Hassman and Baczek18
• Distortion product otoacoustic emission (DPOAE) detection and measurement are influenced by cochlear and middle-ear status
• The effect of chronic suppurative otitis media (CSOM) on DPOAE input–output functions in conventional and ultra-high frequencies was investigated
• The presence of CSOM significantly reduced DPOAE amplitude, especially in conventional frequencies
• The presence of CSOM reduced the rate of ears with measurable DPOAEs in conventional frequencies
• The DPOAEs at ultra-high frequencies may be used to estimate cochlear function in CSOM individuals
• The DPOAE input–output function may be preferred over distortion product gram for CSOM, as it provides more information
The rate of measurable DPOAEs increased in the clinical group as the intensity of the pure tones increased from 55 to 70 dB SPL, in all the tested frequencies. The present study finds support from the study by Yeo et al., which investigated the effect of middle-ear effusion on the rate of measurable DPOAEs in children.Reference Yeo, Park, Park and Suh6 The author reported an increase in the rate of measurable DPOAEs with an increase in the primary tone levels. Significant differences between the control and CSOM groups were found at 2 and 3 kHz for 45 and 55 dB SPL primary tone levels. The present study also found a significant difference in the L1 intensity of: 65 and 70 dB SPL at 0.5 kHz, 55–70 dB SPL at 1, 2 and 4 kHz; 65 and 70 dB SPL at 8 kHz; and 60 dB SPL at 14 kHz. The use of higher primary tones was found to increase the rate of measurable DPOAE responses in individuals with middle-ear disorders, similar to the findings of Yeo et al.Reference Yeo, Park, Park and Suh6 Use of DPOAE input–output function may thus be preferred over a routine DP-gram while measuring DPOAEs in individuals with CSOM. Hence, DPOAE input–output function is an effective tool in assessing middle-ear status before and after medical or surgical intervention. The DPOAE input–output function can also be used as a tool for determining the prognostic value of an intervention.
A reduction in DPOAE parameters in individuals with otitis media could be attributed to the fact that the middle ear itself is acting as a barrier for sound transmission in the conventional frequency range.Reference Kei, Brazel, Crebbin, Richards and Willeston11
An animal model study reported a significant difference in DPOAE growth behaviour in individuals with middle-ear dysfunction and cochlear dysfunction.Reference Gehr, Janssen, Michaelis, Deingruber and Lamm19 Distortion product OAE growth behaviour was not affected in middle-ear dysfunction cases, whereas steepened DPOAE input–output functions were observed in cochlear impairment cases, reflecting the loss of compression of the cochlear amplifier. Future studies are required to understand the DPOAE growth function in individuals with CSOM for conventional and ultra-high frequencies, which could help to differentiate middle-ear disorders from cochlear disorders.
Conclusion
This study attempted to determine the effect of CSOM on DPOAE input–output functions in conventional and ultra-high frequencies. There was a significant difference in DPOAE amplitude between the control and clinical groups, especially for the conventional frequency range. In conventional frequencies, the mean signal-to-noise ratios were significantly lower in the clinical group compared to the control group. In the ultra-high frequency range, the DPOAE signal-to-noise ratios were fairly similar across the groups. The significant differences were observed in DPOAE amplitude and signal-to-noise ratio for the f1 intensity level of 55 dB SPL and above. The presence of middle-ear effusion in the clinical group significantly reduced the rate of ears with measurable DPOAEs, especially in the conventional frequency range. Distortion product OAEs at ultra-high frequencies may be used for a rough estimation of cochlear function in individuals with CSOM. Use of the DPOAE input–output function may be preferred over a routine DP-gram, as the former provides more information than the later.
Competing interests
None declared
Appendix 1. DPOAE signal-to-noise ratios
Conventional frequencies
Ultra-high frequencies
Appendix 2. DPOAE amplitude
Conventional frequencies
Ultra-high frequencies
Appendix 3. Screenshots of DPOAE recordings at ultra-high frequencies in the clinical group
